← Atlas Academy · Human Anatomy studio ↗
ACADEMY-002 VISUAL UPGRADE · LOCAL OWNER REVIEW
Human Anatomy prototype visually approved by JR. Integrated Academy experience and remaining faculty visuals await final acceptance. Academic approval of teaching materials is separate.
Eight studios use real segmented anatomical geometry from BodyParts3D 4.0, official IS-A 99% polygon-reduced OBJ distribution. The retained upstream ZIP is 142,903,898 bytes; SHA-256 40665852c49f218326590e204db91064a1ecfc3c6f8cbd7bbbcaac62c7cd409e. Archive and metadata were downloaded directly from the official Database Center for Life Science archive. Its hash records the downloaded bytes; no separate publisher checksum was available for comparison.
The official licensing page, updated 27 February 2025, specifies CC BY 4.0. The exact downloaded ZIP has no embedded contrary license notice. We use the official distribution rather than a differently licensed mirror. Required attribution: BodyParts3D, © The Database Center for Life Science licensed under CC Attribution 4.0 International.
The machine-readable asset manifest records every included source OBJ filename, SHA-256, canonical label, FMA/BP identifier, geometry encoding, binary hash and source input hashes. Original protected University/Method/Core educational files are not included or modified by the visual asset pipeline.
Atlas triangulates OBJ polygons, quantizes positions to unsigned 16-bit coordinates within each mesh’s original bounds, retains topology and selects a safe index width. Normals are recomputed for lighting; rigid orientation and uniform scale adapt the model to the viewer. The original Anatomy set’s maximum quantization error is under 0.0003% of its source body height. No invented anatomy or additional mesh decimation was introduced. Geometry size and SHA-256 are verified before rendering. Camera framing adapts to viewport proportions. Model colours, translucency and separated/exploded views are presentation choices rather than tissue properties.
| Faculty | Visual and interaction | Scientific boundary |
|---|---|---|
| Human Anatomy | 606 source meshes: 403 muscle/component and 203 bone meshes. Muscular, skeletal and combined layers; rotate, zoom, pick, search, isolate, hide and focus. | One adult male reference. Mesh counts are not counts of unique anatomical structures. Source segmentation is approximate. |
| Biomechanics | Four source-labelled leg bones; controllable knee hinge and optional directional vectors. | Pivot approximated from bone bounds. Rigid illustrative motion, not measured human kinematics or computed forces. |
| Exercise Science | Detailed skeletal muscle fibre plate; source-labelled markers, structure list and selected-region magnification. | Two-dimensional illustration, not a simulation of adaptation or energy turnover. |
| Applied Coaching | Three source-labelled arm bones; user-controlled illustrative elbow flexion, observation and camera comparison. | Not motion capture, a technique assessment, or a validated coaching prescription. |
| Nutrition Science | 63 source-labelled digestive meshes; organ selection, isolation, focus and spatial inspection. | Anatomical introduction only. Metabolism and nutrient absorption are not simulated. |
| Recovery Science | 51 brain-region meshes with transparent context revealing the labelled hypothalamus and pineal body. | Not a sleep-stage, recovery or circadian simulation; source structures only. |
| Rehabilitation Science | Focused knee/patella inspection, four source bone meshes, optional illustrative hinge movement. | No cartilage/ligament contact model, pathological interpretation, clinical range or patient data. |
| Men’s Health | Nine reproductive meshes with three translucent pelvic-context bone meshes; source-labelled exploration. | Limited reference anatomy, not comprehensive men’s health content. |
| Women’s Health | Detailed female reproductive plate; source-labelled structure markers and magnification. | Uses its own female illustration; no male reference relabelled as female anatomy. Not rotatable 3D. |
| Psychology & Behaviour | 51 source-labelled brain-region meshes, cortical inspection and hemisphere separation. | Display separation is artificial. No functional neural network or behavioural inference is claimed. |
| Coaching Intelligence | Interactive eleven-faculty knowledge map, prerequisites and forward relationships read from the curriculum database. | Spatial layout is illustrative. No generated academic relationship, automated coaching decision or student progression policy. |
Blausen 0801 SkeletalMuscle.png and Blausen 0399 FemaleReproSystem 01.png are published by BruceBlaus / Blausen Medical Communications under CC BY 3.0. The skeletal muscle file’s current source revision was cropped by CFCF in 2017. Citation: Blausen.com staff (2014), “Medical gallery of Blausen Medical 2014,” WikiJournal of Medicine 1(2), doi:10.15347/wjm/2014.010.
The full downloaded PNGs remain unmodified, including their original labels. Atlas adds source-based selection markers, CSS magnification and optimized WebP preview crops/resizing. These display adaptations are not endorsed by the creators. Exact originals and SHA-256 values are recorded in the manifest. Markers were checked against the visible source labels; no independent academic/clinical validation is claimed.
The viewer self-hosts Three.js 0.180.0 and OrbitControls under the MIT license. No host installation or package dependency addition was required. All anatomical assets, viewer modules and previews are local. No paid assets, SciePro content, production services or cloud resources are used.
Homepage and faculty introductions use lightweight WebP renders. The heavy viewer and geometry load only after Launch interactive studio. Individual subjects load only their selected geometry; full Anatomy compressed geometry is approximately 21.24 MB. That remains a meaningful mobile/slow-network payload, and this upgrade does not claim production network qualification. Detailed layer delivery is gzip-compressed, version-addressed by source digest, and bounded to an explicit geometry name allowlist.
The original isolated prototype measured approximately 2.0 seconds desktop and 1.2 seconds at mobile viewport size over local loopback, with about 61 browser animation frames per second during rotation. These are host-specific measurements, not physical-phone or GPU benchmarks. Integrated measurements, screenshots and failures are retained in Academy’s docs/science-upgrade/ acceptance evidence.
Controls support keyboard operation, named structure lists, visible focus, touch drag/pinch, and explicit rotation/zoom buttons. Rendering is requested only when needed, with no automatic rotation loop. Reduced-motion preferences suppress visual transitions. In a browser without WebGL, loading stops, a real model reference preview appears and guidance links to the standalone studio address for opening in an external compatible browser. The tested external browser is Microsoft Edge. A fallback is not equivalent to interactive 3D.
Source licensing permits reuse; it does not establish scientific teaching approval. These studios are visual introductions, not clinical tools or published courses. They do not change curriculum records, authentication, editorial roles, publication policy or BIOM-001’s unpublished draft state. No enrollment, examinations, certificates, payments, production deployment or ACADEMY-003 functionality is introduced. Final owner visual acceptance and subsequent academic validation remain separate gates.
